Adaptive equalization system

Abstract

An adaptive equalization system that adjusts the spectral shape of a speech signal based on an intelligibility measurement of the speech signal may improve the intelligibility of the output speech signal. Such an adaptive equalization system may include a speech intelligibility measurement module, a spectral shape adjustment module, and an adaptive equalization module. The speech intelligibility measurement module is configured to calculate a speech intelligibility measurement of a speech signal. The spectral shape adjustment module is configured to generate a weighted long-term speech curve based on a first predetermined long-term average speech curve, a second predetermined long-term average speech curve, and the speech intelligibility measurement. The adaptive equalization module is configured to adapt equalization coefficients for the speech signal based on the weighted long-term speech curve.

What is claimed is: 
     
       1. An adaptive equalization method comprising:
 generating a speech intelligibility value from a speech signal by a computer processor; 
 selecting a plurality of predetermined long-term average speech curves; 
 generating a weighted long-term target spectral shape curve by the computer processor based on the plurality of long-term average speech curves and the speech intelligibility value; and 
 adapting equalization coefficients for the speech signal by the computer processor based on the weighted long-term target spectral shape curve; 
 where the plurality of long-term average speech curves change based on speech signal conditions. 
 
     
     
       2. The method of  claim 1  where the plurality of predetermined long-term average speech curves comprise long-term average speech shapes. 
     
     
       3. The method of  claim 1  where the plurality of predetermined long-term average speech curves are selected based on a source of the speech signal. 
     
     
       4. The method of  claim 1  where the plurality of predetermined long-term average speech curves are selected based on gender. 
     
     
       5. The method of  claim 1  where the speech signal conditions reflect quiet conditions or noisy conditions. 
     
     
       6. The method of  claim 1  where the plurality of predetermined long-term average speech curves are selected based on a level of vocal effort used to generate the speech signal. 
     
     
       7. The method of  claim 6  where the level of vocal effort reflects a normal vocal effort, a raised vocal effort, or a shout vocal effort. 
     
     
       8. The method of  claim 1  further comprising adjusting one or more of the plurality of predetermined long-term average speech curves based on a user's environment. 
     
     
       9. The method of  claim 1  further comprising adjusting one or more of the plurality of predetermined long-term average speech curves based on an additive noise level. 
     
     
       10. The method of  claim 1  further comprising adjusting one or more of the plurality of predetermined long-term average speech curves based on room acoustics. 
     
     
       11. The method of  claim 1  further comprising adjusting one or more of the plurality of predetermined long-term average speech curves based on a microphone frequency response. 
     
     
       12. The method of  claim 1  where the weighted long-term target spectral shape curve comprises a template. 
     
     
       13. The method of  claim 12  where the template comprises an output template. 
     
     
       14. The method of  claim 1  further comprising adjusting the equalization coefficients in response to signal conditions. 
     
     
       15. The method of  claim 1 , where generating the speech intelligibility value comprises:
 obtaining a signal power measurement for a frequency band of the speech signal; 
 obtaining a background noise level for the frequency band of the speech signal; and 
 obtaining the speech intelligibility value from the signal power measurement and a background noise level associated with the frequency band of the speech signal. 
 
     
     
       16. The method of  claim 1 , where adapting the equalization coefficients comprises:
 applying a prior version of the equalization coefficients to a power spectrum of the speech signal to generate an equalized signal; and 
 adapting the equalization coefficients to generate an adapted version of the equalization coefficients based on the equalized signal and the weighted long-term target spectral shape curve. 
 
     
     
       17. An adaptive equalization system, comprising:
 a computer processor; 
 a speech intelligibility measurement module executable by the computer processor to process a speech intelligibility value of a speech signal; 
 a spectral shape adjustment module executable by the computer processor to process a weighted long-term target spectral shape curve that varies with speech signal conditions and is based on a plurality of predetermined long-term average speech curves and the speech intelligibility value; and 
 an adaptive equalization module executable by the computer processor to process equalization coefficients for the speech signal based on the weighted long-term target spectral shape curve. 
 
     
     
       18. The system of  claim 17  where the plurality of predetermined long-term average speech curves include a first speech template in quiet conditions and a second speech template in noisy conditions. 
     
     
       19. The system of  claim 17  where the plurality of predetermined long-term average speech curves are selected based on gender. 
     
     
       20. The system of  claim 17  where the plurality of predetermined long-term average speech curves are selected based on a level of vocal effort that reflects a normal vocal effort, a raised vocal effort, or a shout vocal effort.